1. Field of Invention
This invention relates to electronically-controlled fuel injection valves and particularly to fuel injection valves of the type having a stem-mounted ball-valve telescoped by the front end and secured to the back end of a tubular armature having a flow smoothing passage therethrough.
2. Description of Prior Art
Conventional fuel injection valves, such as of the type disclosed in the patent to Kirsch U.S. Pat. No. 3,828,247, comprise one of the most expensive components of fuel injection systems in current mass production for passenger vehicles. Such conventional injectors incur such comparatively high costs because most of the structural elements effecting fuel breakup, fuel spray angle, fuel metering, and flow on/off valving are made to extremely close tolerances. Meeting these tolerances requires specialized lapping by a tool that cannot be used again for final lapping, and the resulting parts are custom rather than randomly mated. Even then such conventional fuel injection valves do not normally breakup the fuel into controllably uniformly small particles and thereby limit the attainment of both maximum fuel economy and minimum formation of undesirable emissions. Moreover, comprising extremely-narrow and closely-toleranced fuel metering and breakup paths, such conventional valves are susceptible to the deleterious effects of contamination passing the inlet filters of the injectors or back flowing from engine inlet passages into the injector outlet sections. It is therefore desirable to reduce the cost of fuel injection valves by avoiding the conventional lapping and redressing, custom hand mating, and generally tight tolerancing all over.
A primary factor imposing harsh tolerancing requirements on such conventional fuel injection valves is the use of different elements of just one part, a reciprocating pintle-type needle-valve member, to perform the breakup, metering, and valving functions. Each such different element must be closely concentric not only with the other elements of the same part but also with each of the surrounding structures cooperating with such elements.
The present invention recognizes that at least the close concentricity tolerances could be substantially relaxed and in turn other gross cost savings obtained by effecting the on/off valving function by a structure substantially separate from that effecting the fuel breakup function and the metering function. More specifically, recognizing that a circular seating edge need not be closely concentric with the metering orifice, the invention allows the use of no less than three cost saving processes: (1) the conventional loose-concentricity low-cost "ballizing" process of forcing a final diameter precision ball through a initially undersized aperture to repeatedly provide highly finished uniform orifices; (2) the conventional loose-concentricity low-cost "coining" process of forcing a precision ball against a softer conical surface to repeatably provide a circular nonleaking seating edge; and (3) the conventional loose-tolerance ball valve head and oversized ball seat technique to repeatably effect the on/off valving. Thus, even though the patents to Mattson U.S. Pat. No. 1,360,558 and Seccombe U.S. Pat. No. 3,587,269 suggests the use of ballizing, even though the patent to Carlson U.S. Pat. No. 3,400,440 suggests a fuel injection valve having a ball seat coined by a slightly larger ball, and even though the patent to Malec U.S. Pat. No. 3,490,701 suggests the use of a stem-mounted ball valve, such prior art not withstanding fuel injection valves are not known to have heretofore used any combination of a ballized metering orifice, with a coined valve seat, or a stem-mounted ball valve head, perhaps because of the severe concentricity requirements previously thought to be essential. As indicated above, a primary function of a fuel injection valve is to break up a metered quantity of fuel into combustible particles. Generally, the smaller the fuel droplets, the more readily they vaporize for combustion and the more completely they burn. Moreover, the more complete and efficient the combustion, the better the brake specific fuel consumption or mileage and the less the generation and emission of undesirable exhaust emissions. Conventional injectors of the type disclosed in the above-mentioned Kirsch patent develop a spray by forcing fluid between a closely toleranced needle and its single surrounding closely-toleranced annular orifice, and the resulting drop sizes comprising such spray are of varied sizes and distributions depending on the actual dimensions of the annular orifice. Moreover, while a fuel injector using a plurality of circular apertures through a thick plate is disclosed in the patent to Harper Jr. U.S. Pat. No. 2,382,151, such circular apertures generate a generally pearshaped solid cloud of fuel particles rather than control the size or variation thereof of the particles. Moreover, circular holes of the requisite smallness are difficult to fabricate repeatably even by etching. The analysis by Rayleigh in his "On the Capillary Phenomena of Jets" (Proceedings of the Royal Society, XXIX pp 71-97, 1879, Rayleigh, Scientific Papers, Vol. 1, Dover Publications, 1964) is therefore also of interest to the present invention. There Raleigh noted that non-circular orifices through thin plates produced flat broad thin liquid sheets of fluid. More recent analyses, such as by Keller and Koldner in the Journal of Applied Physics Vol 25 pp 918-21 (1954), show that thin sheets produce small droplets. However, it was not appreciated until recongized by the present invention that non circular slots of the requisite small width could be etched more precisely than circular apertures with the result that the thin-plate-non-circular-slot thin-liquid-sheet small-droplet theory is not known to have heretofore been applied to fuel injection valves. It is therefore desirable to improve fuel economy while at the same time reducing undesirable emissions by breaking up the metered fuel first into thin sheets and then into uniformly small fuel droplets. Conventional injection valves of the type noted above do little if anything to shape the envelope of the spray emitted from the annular orifice. This results in a wide angle spray that wets the sides of the intake passages so as to enter the combustion chamber in an unevenly rich and lean distribution. The present invention recognizes that such wetting and uneven distribution may be reduced by providing a spray-envelope-shaping nozzle as a part of the injector immediately downstream of the fuel breakup disc.
The pressure drop across the fuel breakup means of a conventional fuel injection valve is another factor requiring very tight tolerancing of not only the metering orifice but also the breakup apertures. Since the precision of the quantity of fuel injected on each injection pulse is dependent on having a known flow rate while the injection valve is open and since a known flow requires having a known pressure drop across a known flow area, the area of any part of the flow path across which there is any significant pressure drop must be known and therefore closely controlled. The present invention therefore further recognizes that the size tolerances on the fuel breakup means could be relaxed by effecting the breakup function by a structure substantially separate from that effecting the metering function and by then designing the fuel breakup means so as to have a minimum pressure drop thereacross. In otherwords, the present invention recognizes the desirability of providing fuel breakup means having a sufficient flow area and minimal axial thickness so as to not generate any pressure drop significant to fuel flow accuracy. In this way the tolerances on the non-circular breakup apertures could be determined, not so as to effect a requisite pressure drop by means of a precisely known flow area therethrough, but rather to effect the requisite drop size, the tolerances on the breakup apertures being looser than those on a metering orifice. Moreover, the tolerances on the breakup apertures could then be held by the low cost etching through thin plates.
Conventional fuel injection valves introduce an undesirable, and often vehicle disabling, "hot start" problem upon restarting or attempting to restart an overly hot engine before it has had sufficient time to cool down. More specifically, during the comparatively short time between shutting down an engine in an overly hot environment and attempting to restart the engine, all the components under the hood experience a "hot soak" as the overly hot engine conducts, convects, and radiates heat to the auxiliary components. In the case of the fuel injection valves, the temperatures thereof are so elevated compared to the temperatures associated with normal operation that the fuel is substantially vaporized before reaching the valving and metering elements. To the extent that the fuel is vaporized prior to being metered, less liquid fuel is expelled from the injector during a given injection interval than is expelled under normal operating conditions when the fuel is substantially liquid. Consequently, to the extent that more vaporized than liquid fuel is injected into the inlet passages of the engine, a substantially leaner than desired mixture is injected. Such leaner mixture is often insufficient to permit proper ignition, preventing ignition under the worst cases and otherwise effecting stumbling to rough ignition under less severe cases as the mixtures richen up to the desired air-fuel ratio. The duration of such undesirable lean mixture performance varies primarily with the difference between the hot soak and normal operating temperature and the rate at which the hot soak thermal energy is removed from the injector.
To avoid such "hot restart" problems, it is desirable to reduce the problem-causing conduction, convection, and radiation of heat from the engine to the injectors and then to eliminate whatever hot soak energy is transfered thereto as fast as possible upon hot restarting. More specifically, it is desirable to minimize the initial conduction of hot soak energy to the injectors by minimizing the surface contact area between the engine and the injectors and by minimizing convection and radiation by increasing the air space between the exterior of the engine and the exterior of the injector. Furthermore, to reduce the time required to remove whatever heat has been transfered to the injectors, it is desirable to reduce the cross-sectional area of the injectors so as to increase the air space between the engine and injector, to reduce the stored hot soak energy that must subsequently be removed, and to otherwise maximize the rate that heat is transferred from the body of the injectors.
In solving this problem, the present invention recognizes that smoothly-flowing normally-cooler fuel has a higher coefficient of heat transfer than turbulently flowing fuel and, not being turbulent, can be metered more precisely. In this regard, the present invention recognizes that it is desirable to induce a substantially smooth flow and to do so by a substantially straight and unimpeded central fuel flow immediately upstream of the valve and orifice rather than the prior art side-ported and peripherally-chanelled fuel flow of the types produced by the valves disclosed in the above mentioned patents.
A further primary function effected by a fuel injection valve is to repeatably and rapidly actuate the valve by the electromagnetic interaction between the flux produced by a fixed coil acting on a movable plunger or armature connected to the valve head. Conventionally, the actuator is electromagnetically opened to a position determined by the abutment of a shoulder protruding from the actuator against suitable abutment on the valve body such abutment normally being in the form of a "C" washer. Upon de-energization of the coil the actuator is spring closed to a closed position determined by seating of the valve head on the valve seat. To effect as rapid a response as possible with the establishment of a threshold level of magnetomotive force by the coil, the actuator is made as light as possible and the magnetic lock up between the fixed and movable elements is prevented by maintaining minimum magnetic air gaps for the magnetic flux. In addition to permitting a faster opening response, a light actuator permits the use of a weaker closing spring to effect softer closing and thereby also reducing the pounding wear between the valve head and valve seat. The outer surface of a conventional actuator and the mating inner surface of a conventional actuator housing are therefore heat treated and closely toleranced as to diameter and squareness so as to provide a durable sliding metal-to-metal contact. Such close tolerancing is required: (1) to enable the actuator to precisely pilot and center the valve head on the valve seat; (2) to precisely pilot and center the pintle needle in the metering orifice; and (3) to maintain the minimum magnetic air gaps axially between the rear end of the armature and the front of the fuel inlet tube and also radially between the outer diameter of the armature and the inner diameter of the mating valve body. It is desirable to avoid heat treatment and relax these tolerances especially since they must otherwise be maintained on substantially blind and very small actuator housing bores.
The present invention recognizes that an actuator which is tubular in form enhances such lightness in addition to also inducing a smoothing better-cooling-and-metering effect on the central flow therethrough. Moreover, the present invention further recognizes that, rather than providing a sliding metal-to-metal contact between the actuator and its housing, it is more desirable to do the opposite by providing an ample positive clearance therebetween to allow the resulting surrounding pressurized fluid fuel to sufficiently center the actuator to effect the necessary seating and to maintain the minimum air gaps. Also, lower actuation energy is required when the actuator slides on a fluid rather than metal surface, also permitting a weaker closing spring resulting in lower closing impact and longer actuator life. The present invention further recognizes that a positive clearance between the actuator and its housing also enables the actuator to provide some of the flexing action otherwise required of the stem to properly seat the stem-mounted ball valve head on the valve seat. More specifically, the length of the actuator telescoping the stem and free to move in the positive clearance acts as extension of the stem and thereby reduces the life limiting flex stresses that would otherwise be imposed thereon.
A further cost imposing feature of conventional fuel injection valves heretofore used with commercial passenger vehicle fuel injection systems is that the electromagnetically responsive armature is mounted on a non-magnetic actuator. Not only is the non-magnetic material more costly per pound by half again as much as the magnetic material, but the separate armature and actuator parts require close tolerance machining of the requisite mating concentric bores in the armature and receiving surfaces on the actuator followed by the close tolerance axial positioning of the armature on the actuator. The main reason requiring such separate materials apparently was the previous belief that, unless the actuator was of non-magnetic material, the motion limiting stop shoulder thereof would effect a magnetic lock-up with the magnetic return path of the valve body and would thereby unacceptably slow the opening and closing times of the injector.
The present invention recognizes that any magnetic lock-up between the actuator shoulder and valve body is second order compared to that possible between cylindrical outer surface of the armature and valve body because the latter provides not only the shorter flux return path inherently effected by magnetic flux but also provides more mating gap surface. The present further recognizes that, rather than suffering the cost and other penalties of providing an armature and actuator of different materials, it is feasible and more desirable to do the opposite by making not just the armature and actuator but also the actuator housing out of the same material. By doing so avoids the differential thermal expansion rates heretofore resulting from different coefficients of expansion. Also avoided is the growth of crystals in the gaps normally resulting from the galvanic corrosion reaction conventionally occuring between the dissimilar materials of the actuator and its housing, such similar material thereby further reducing the friction therebetween while increasing valve life by avoiding catastrophic galvanic-growth-induced seizure of the actuator to its housing.
Yet another problem heretofore experienced with electromagnetically actuated fuel injection valves is that the welded connections between the end of coil wire and the output terminal of the injector often break when the output terminals are wiggled on the assembly, connector molding, testing, shipping, or subsequent engine mounting and connection of the injector. Conventional fuel injection valves of the type noted above attempt to avoid these problems by the use of L-shaped terminals that enter the injector axially and then, make an "L" turn in opposing circumferential directions so that the inside of coil bobbin and/or inlet connector flange prevents the terminals from being moved axially. Such terminals of course are not stamped out from lower cost straight ribbon stock of terminal width. It is therefore desirable to provide a straight narrow terminal that can be securely anchored within the bobbin.